The Salesforce IoT Cloud

There's nothing like hitting all of the trends with one product. This year, the Internet of Things (IoT) is hot, hot, hot, as is Big Data. Roll in analytics, and you have the trifecta of trendiness.

Salesforce is running that race with its IoT Cloud.

Announced at the company's annual user conference, Dreamforce, last month, and to be available next year, the IoT Cloud arrived with a bang – literally. CEO Marc Benioff, along with Salesforce co-founder Parker Harris, playfully riffed on a Norse mythology theme, complete with Thor's hammer, to show off the IoT Cloud, powered by Thunder, a massively scalable, real-time event processing engine.

Yes, it was noisy fun. But behind it is technology that's designed to let the Things from the IoT interface with Salesforce's more mainstream technology. The IoT Cloud can handle the rapidly streaming volume of data from the Things in the IoT, make sense of it, and direct outputs to Salesforce's other Clouds.

For example, sensors on machinery can, through the IoT Cloud, open tickets in the Service cloud when the equipment needs attention. No human intervention is required. Late flights can trigger automatic rebooking of customers who will miss their connections. Customers can be notified on their mobile phones if a device is operating outside of acceptable parameters, for example, if the kids manage to kick the thermostat at home to an extra-high temperature.

Internet of CustomersIn addition to connections to the "traditional" Internet of Things, such as phones, wearables, windmills, industrial turbines, and other devices, IoT Cloud connects data from websites, blogs, social interactions and more to Salesforce, to bring customer context to transactional data. Salesforce refers to this melange as the Internet of Customers, based on the premise that behind every "Thing" is a customer.

IoT Cloud lets users create rules and logic for events coming from all of these inputs, which in turn create actions in other components of Salesforce's clouds. A sensor that indicates a vehicle is being driven aggressively could trigger an alert to head office that there may be maintenance needed (and possibly a firm chat with the offending driver). A retailer having a holiday sale could use beacon technology to send special offers to members of its loyalty program when they walk into the store. Weather forecasts, sensor readings, and temperature settings could be combined send advice to an HVAC system's owners on how best to optimize its use.

In fact, use cases are only limited by the customer's imagination. If something has a way to input to the IoT Cloud (Salesforce's API lets vendors make those connections; initial launch partners will be ARM, Etherios, Informatica, PTC ThingWorx and Xively LogMeln, providing connectivity between devices and the Internet.), and there's a useful function that could be performed by one of Salesforce's clouds, the IoT Cloud will enable it. Want to launch a sales initiative for a new car when maintenance issues on the old one reach a certain threshold? Need to contact customers expressing negative sentiment about your brand to salvage the situation, or send reminders to customers to make a service appointment based on telematics? All do-able.

There are, of course, caveats. While the Things connecting with systems can provide benefits, if they're attached to personally identifiable information (PII), they generate privacy issues. Those have to be addressed, both for legal and compliance reasons, and to reduce the creepiness factor.

And it can get creepy. One story that quickly went viral (and is now almost a legend) involved a young woman who began receiving promotions from Target for baby products, before she had told her family of her situation. It did not go well.

Today, companies have learned that they shouldn't be too blatant with their electronic stalking – the baby ads would appear mixed in with other ads, in a relatively unobtrusive spot in a flyer, not front and centre.

Companies have also learned about ways to "stalk" customers productively for all concerned. For example, with the customer's consent, in-store beacons can identify shoppers, determine their preferences through shopping history, and guide them to deals that they might find attractive, all through an app on their phones. One clothing store goes so far as to light up shelves containing items that would co-ordinate with previous purchases to guide customers to them.

Sensors within a store, combined with a phone app, can track a customer's path and prompt a sales associate to assist. Analytics that analyse patterns, combined with the sensors detecting them, deduce what the customer's goal may be. If, for example, someone wandering through a hardware store checks out vanities and sinks, he might be guided to the faucets and provided with a sales associate to assist in designing that bathroom he may be renovating.

Salesforce's IoT Cloud, of course, is designed to tie into Salesforce's other products. Things can commune with the Sales Cloud, the Service Cloud, the Marketing Cloud, and even with apps. But it will also talk to third party products, for example, to email a customer whose monitored smoke detector's battery is getting low.

The big thing about this technology is its relative simplicity. There's no programming involved in setting rules to trigger actions, or to reporting on sensors. It's meant to allow almost anyone to interact with the Things in their world, and to generate useful actions or reports with minimal fuss.

The IoT Cloud will enter pilot testing in early 2016, and become generally available later in the year. Even Salesforce has no idea what people will do with it; when quizzed about the devices it will support, EVP Tod Nielsen said it would depend on what customers ask for.

This is going to be interesting to watch.